Server computers (or servers) provide data files, and other resources, to client computers (or clients) over a network. Servers present a set of resource or file names, called a namespace, in which all the names are unique, whereby the clients can access shared resources provided by the servers. Servers are accessed by clients using a server name or other server identifier. Over time, and as network configurations change, a server can become identifiable by more than one server name. As a result, server name resolution presents challenges for accessing resources on a network.
For example, old servers are often updated by replacing them with newer, faster servers. Unfortunately, when servers are consolidated, the namespaces can change such that namespaces provided by the old servers are no longer useable by clients of the new servers. Thus, server consolidation can render resources inaccessible to clients, unless the clients are updated to recognize the new namespaces. Updating the clients with new server names can be a time consuming, expensive task that can disrupt work flow at the clients. In addition, consolidation of servers can cause collisions among share names that are common among the consolidated servers.
When multiple servers are consolidated into fewer servers, a resource that was available on one of the old servers, but not available on the other old servers, may become available to clients of the other old servers through one of the new servers. Clients of the new servers may not know that new resources are available because of the consolidation. Even if the clients know of the new resources, it may not be apparent how to access the new resources. Thus, consolidating servers may have benefits, such as economies of scope, which are not fully realized by current systems.
In addition, a server can be identified by server names as well as an internet protocol (IP) address. In some configurations, a server name is mapped to an IP address by a Domain Name System (DNS) server. However, someone who knows the IP address of a server can access the server directly. Existing legacy software may include ‘hard-coded’ IP addresses to servers. In addition, some clients may use a Network Basic Input/Output System (NetBIOS) server naming convention that differs from a DNS server name. When server configurations change, as, for example, when a DNS server is installed, it may no longer be desirable or possible to directly access a network server using a previous IP address or a NetBIOS name. Unfortunately, identifying and updating all the clients and software that use NetBIOS names and IP addresses can be impractical.